Telegraf AI Prompts
27 free, production-ready Telegraf AI prompts — copy-paste ready for ChatGPT, Claude, or Cursor. Every prompt ships with fill-in placeholders, a worked example, and production-safety notes, so you get a usable answer on the first try.
- Telegraf Advanced
Control High Cardinality in a Telegraf Pipeline
Find and cut cardinality-driving tags in a Telegraf pipeline using processors (regex, enum, converter, dedup), tag stripping, and aggregation so storage stays healthy and queries stay fast without losing the dimensions that matter.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Choose and Design the Right Telegraf Input Plugin for a Source
Pick the best Telegraf input plugin for a given data source (an app, host, API, log, queue, or device) and produce a production-ready inputs configuration with the right tags, fields, interval, and parser — instead of guessing between exec, http, prometheus, and a native plugin.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Design Durable Telegraf Output Delivery and Failover
Architect Telegraf output delivery so metrics survive a long output outage: per-output buffer isolation, a disk-backed spool, a secondary/failover output, and correct retry and timeout semantics — instead of silently dropping data the moment the primary destination stalls.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Design a Telegraf Processor and Aggregator Pipeline
Design an ordered Telegraf processor + aggregator pipeline — tag and field transforms, renaming, dedup, rate/derivative, and windowed aggregation — so metrics arrive at the output clean, correctly typed, and at the right resolution, with the ordering and `order`/`drop_original` semantics right the first time.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Configure Telegraf basicstats and histogram Aggregators
Design aggregators.basicstats and aggregators.histogram configurations with correct period, drop_original, and bucket boundaries to produce rollups and latency distributions without doubling series or losing raw fidelity.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Telegraf Docker Input Metrics Collection Prompt
Configure inputs.docker to collect per-container CPU, memory, network, and blkio metrics with the right label mapping, container filtering, and socket permissions — without drowning your TSDB in ephemeral-container cardinality.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Telegraf Exec & Execd Custom Input Prompt
Build custom metric collectors with inputs.exec (run a script per interval) and inputs.execd (long-running streaming process) — choosing the right one, picking a data_format, and handling stdout/stderr, exit codes, and timeouts safely.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Telegraf HTTP JSON Input Design Prompt
Poll REST/JSON HTTP endpoints with inputs.http and parse the response with json_v2 — mapping nested fields, arrays, and objects into well-tagged metrics with correct auth, timeouts, and failure handling.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Design a Telegraf outputs.influxdb_v2 Output for Reliable Writes
Configure the outputs.influxdb_v2 plugin with correct org/bucket/token handling, batching, retries, timeouts, and content-coding so writes stay durable and efficient against InfluxDB 2.x / Cloud without dropping metrics under backpressure.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Telegraf IPMI Hardware Metrics Prompt
Collect bare-metal health with inputs.ipmi_sensor — temperatures, fan speeds, voltages, PSU and chassis status from BMCs over local or LAN interfaces — with safe credentials, sane intervals, and tag mapping for fleet dashboards.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Telegraf Jolokia JMX Input Prompt
Collect JVM and application MBeans through Jolokia with inputs.jolokia2_agent — heap/GC, thread pools, and app-specific metrics — mapping MBean attributes to tags and fields with the right auth, batching, and cardinality control.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Telegraf Kafka Consumer Input Prompt
Configure inputs.kafka_consumer to ingest metrics/events from Kafka topics — consumer group balancing, offset handling, TLS/SASL auth, message parsing, and backpressure — so Telegraf becomes a reliable stream consumer, not a lag generator.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Design a Telegraf outputs.kafka Pipeline with Topic Routing
Configure the outputs.kafka plugin with dynamic topic routing, partition keys, serialization, compression, acks, and TLS/SASL so Telegraf publishes metrics to Kafka reliably and with sane partition distribution.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Deploy Telegraf as a Kubernetes DaemonSet
Design a Telegraf DaemonSet (plus optional Deployment for cluster-scoped metrics) with correct RBAC, resource limits, env/secret injection, config via ConfigMap, and node/pod scoping so metrics collection is complete but not duplicated or resource-abusive.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Tune Telegraf Metric Buffer, Batch, and Flush Settings
Right-size metric_buffer_limit, metric_batch_size, flush_interval, flush_jitter, and collection_jitter across the agent and outputs so Telegraf survives output outages and traffic spikes without dropping metrics or thundering-herd flushing.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Beginner
Telegraf Net Response & Port Check Prompt
Build synthetic reachability and latency checks with inputs.net_response and inputs.ping — TCP/UDP port probes, expected-string matching, timeouts, and result-code mapping — so connectivity failures become clean, alertable metrics.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Telegraf Procstat Process Monitoring Prompt
Design inputs.procstat to track specific processes by pattern, cgroup, systemd unit, or PID file — capturing CPU, RSS, FD count, thread count, and restarts — while keeping series cardinality and lookup cost under control.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Configure Telegraf outputs.prometheus_client for Scrape Exposure
Stand up the outputs.prometheus_client plugin as a scrape target with correct metric_version, expiration, TLS, and string/collector handling so Prometheus reads stable, correctly-typed metrics without stale or duplicated series.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Telegraf Prometheus Scrape Input Prompt
Use inputs.prometheus to scrape /metrics endpoints and Kubernetes pods — target discovery, relabeling, metric_version selection, and namespace/label filtering — so Telegraf replaces or complements a Prometheus server without cardinality blowups.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Build Telegraf regex and enum Processors for Tag Normalization
Compose processors.regex and processors.enum blocks to normalize noisy tag/field values — rewriting labels, extracting fields from paths, mapping status codes to human states, and collapsing high-variance values — without exploding series cardinality.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Manage Telegraf Credentials with Secret Stores
Move tokens, passwords, and keys out of telegraf.conf into secret-store plugins (os keyring, systemd credentials, HashiCorp Vault, cloud secret managers) using the @{{ secretstore.key }} reference syntax so no credential is ever committed in plaintext.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Telegraf SNMP Input Plugin Design Prompt
Design a scalable inputs.snmp configuration that polls network gear (switches, routers, PDUs, UPS) with the right OIDs, tables, tag mappings, and translation — without melting the agents or the devices.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Telegraf SQL Database Input Prompt
Turn arbitrary SQL queries into metrics with inputs.sql — driver/DSN setup, mapping result columns to tags and fields, query scheduling, and timeouts — so business and DB-internal metrics flow into your TSDB without a custom exporter.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Design a Telegraf Starlark Processor for Metric Transformation
Write, test, and harden a processors.starlark script that reshapes Telegraf metrics — renaming, deriving fields, computing rates from counters, splitting/merging metrics, and dropping noise — with correct handling of the apply(metric) contract and per-instance state.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Migrate StatsD and collectd Ingestion to Telegraf
Plan and execute a migration from StatsD servers and collectd daemons onto Telegraf's inputs.statsd and inputs.collectd (or socket_listener) with parallel-run validation, metric-name/tag mapping, and cutover so no metrics are lost or renamed unexpectedly.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Intermediate
Route Telegraf Metrics with tagpass, namedrop, and Filters
Design metric filtering and routing using tagpass/tagdrop, namepass/namedrop, fieldpass/fielddrop, and taginclude/tagexclude so the right metrics reach the right outputs and processors without accidentally dropping or duplicating streams.
- Claude
- ChatGPT
- Cursor
Open prompt - Telegraf Advanced
Telegraf Tail Input: Logs to Metrics Prompt
Turn plain-text and structured log files into metrics with inputs.tail — grok/regex/JSON parsing, multiline handling, and offset tracking — so error rates and latencies flow into your TSDB without a full log pipeline.
- Claude
- ChatGPT
- Cursor
Open prompt
About these Telegraf AI prompts
This is a free, curated collection of Telegraf AI prompts for DevOps and platform engineers — prompt templates that turn Claude, ChatGPT, or Cursor into a focused Telegraf assistant. Instead of re-explaining your setup every time, each prompt has fill-in placeholders, a realistic worked example, and explicit safety and back-out notes, so it's safe to use against production Telegraf.
Browse all 27 above, or explore the wider Telegraf prompts and troubleshooting guides, the full DevOps AI prompt library, and run or compare any prompt live in the Prompt Workspace. Hit a production issue instead? Try the free Incident Assistant.